• DOI:10.2307/3872535
• Corpus ID: 90555276

Patterns of Weight Gain and Loss for Grizzly Bears in Northern Canada

@inproceedings{Kingsley1983PatternsOW,
  title={Patterns of Weight Gain and Loss for Grizzly Bears in Northern Canada},
  author={Michael C. S. Kingsley and John A. Nagy and Richard H. Russell},
  year={1983}
}

• M. Kingsley, J. Nagy, R. Russell
• Published 1983
• Biology

Seasonal weight change in the grizzly bear (Ursiis arctos) in two populations in northern Canada was studied by fitting growth curves to spring and fall weights. The spring weight of females levels off soon after the average age of first reproduction; that of males continues to increase through maturity and eventually reaches nearly twice that of females. Males lost in winter 22% of their fall weight, the fraction changing very little with age. The weight change in females is much larger, and continues to increase with age, the oldest females gaining in summer 70% of their spring weight and losing in winter 40%/ of their fall weight. Mature females gain and lose not only relatively, but absolutely, more weight than males. Int. Conf. Bear Res. and Manage. 5:174-178 The grizzly bear undergoes an annual cycle in weight, gaining in summer when it is active and feeding, and losing in winter when torpid (Pearson 1975). This weight gain and loss has not, to date, been quantified in free-ranging grizzly bears. Some information has been obtained by comparing weights when the same bear has been fortuitously captured at different times, but such recaptures are few: Pearson (1975) gave only 4 values for winter weight loss, and 2 for rate of gain in fall; Russell et al. (unpubl. rep. for Parks Can., Can. Wildl. Serv., Edmonton, 1979) gave 3 for each. In this paper, seasonal changes in weight have been studied by comparing growth curves fitted to age-weight data for spring captures and for fall captures. Wood et al. (1962) used a similar method, fitting growth curves of the von Bertalanffy type to the seasonal extremes of weight in subspecies of Odocoileus hemionus. The data used here were obtained from studies on the grizzly bear in the Arctic mountains of the northern Yukon Territory, and on the Tuktoyaktuk Peninsula and Richards Island on the coast of the District of Mackenzie, Northwest Territories. These studies were undertaken to establish population status and life history parameters with which to evaluate the effect of hunting, and to provide baseline data for environmental impact assessment. The full reports are in preparation (by Pearson et al. and by Nagy et al.). i Present address: Arctic Management Research, Dept. Fisheries and Oceans, Winnipeg, Man. R3T 2N6 2 Present address: Alberta Environmental Centre, Box 4000, Vegreville, Alta. TOB 4LO METHODS Study areas were selected in the northern Yukon and on the Tuktoyaktuk Peninsula. The latter is topographically homogeneous and of low relief. The northern Yukon study area is topographically more varied and of higher relief. However, grizzly bears have been shown, by radio-collaring, to range widely both here and elsewhere (Pearson 1975; Russell et al., unpubl. rep. for Parks Can., Can. Wildl. Serv., Edmonton, 1979) so the population can be treated as genetically homogeneous. Bears were immobilized by injection of phencyclidine hydrochloride, and weighed to the nearest kilogram by being suspended in a net from the cargo hook of a hovering helicopter. A premolar tooth was taken from each bear and fixed for later decalcification, sectioning, and staining (Mundy and Fuller 1964, Stirling et al. 1977). Ages were determined by counting cementum annulations in sections at lOOx magnification. Bears handled in May and June were assigned ages of n + 0.3 years; in late August and September, n + 0.7 years. Data for bears captured in midsummer were not used in this analysis. A weighted least-squares method (Kingsley 1979) was used to fit von Bertalanffy growth curves of the form W = Woo [1-exp (-kt-x)1]3 where W = weight (kg) at age t (years), Woo = asymptotic weight (kg), k = relative growth rate (per year), and x = fitting parameter; and to derive the standard errors and error correlations of the parameter estimates.